A Kalibr Basin Census Brief

Every compressor set in a basin, counted.

One basin every Wednesday, free in full: who set the iron, who runs it, what engine, what vintage, assembled from the public regulatory record.

Get The Iron Count Bring us your basin

You are negotiating against a number nobody will show you.

Ask your compression vendor what the market rate is and you get a range. Your peers cannot help, because they are working from the same vendor conversations you are, and none of it is knowable from public filings anyway. Customer 10-Ks say "third parties." Vendor 10-Ks report national fleet totals and utilization percentages. Neither side names the other, and the gap between what your vendor knows and what you can verify is the negotiation.

The public regulatory record closes that gap. Every compressor set in a producing state generates a filing: who set it, where, what engine, what horsepower, when. Assembled operator by operator and normalized to the company, it becomes a public map of who runs whose iron.

That map is The Iron Count. It is free, in full, every issue.

The Iron Count sheet masthead: Kalibr Partners, The Iron Count, a Kalibr Basin Census Brief, Permian, No. 01.

FIVE PAGES, EVERY BASINNO MNPI

The citation line prints on every cover: The Iron Count, Permian, No. 01. Numbering is series-wide.

Four things the count settles.

Two layers, both printed

The share of a basin's horsepower the public regulatory record attributes to a named vendor, set beside the standing rented share the vendors report in their own filings. Both are true and they are not the same number. The issue prints both and never blends them.

What counts as a new set

Most new sets are machines that already existed, moved. With big-bore engine lead times where they are, redeployment is where the horsepower actually comes from, and a dated census is how you watch it move.

The renewal wall, with names

Three-year terms are the market standard, and every unit in the census carries a date. So the cohort rolling off in a given half is a list rather than a guess, with the operators named.

The filings audit

Public operators and vendors tested against their own filings, earnings calls, and lease footnotes: ConocoPhillips, EQT, Williams, ONEOK, Kinder Morgan, Energy Transfer, EOG, Archrock, Kodiak, USA Compression, and others. Every conclusion carries an explicit probability. Some hold, others get re-proportioned, and one has been refuted outright. All of it prints.

Each basin returns every quarter on refreshed counts, so the issue also carries what moved: sets added, horsepower added, vendor pairings that changed hands, and where the engine mix is drifting.

The same five pages, every basin.

PAGE 01The market, and the basinThe numbers everyone is negotiating against, set beside what the census holds for that basin.

PAGE 02Basin activitySets per year with horsepower, printed with per-state coverage dates so you can see where the record ends. Sub-basin and county splits. The latest half's setters, ranked.

PAGE 03The ironEngine family census. Build vintage, and what share of the horsepower was built this decade. The dominant engine family for every major operator, because the engine tells you the program.

PAGE 04The counterpartiesThe operator league by horsepower set, each book's vendor attribution, and the leading vendor behind it. The ownership spectrum from pure owners to anchor renters. Then the filings audit.

PAGE 05Five challenges, answeredThe five questions the industry is negotiating blind, each answered with data and each source chipped. Plus the full method: scope, folds, exclusions, coverage dates, suppressions.

Every figure carries its provenance

CENSUSderived from the public regulatory recordVENDOR-STATEDa company's own filingANALYST-ESTa named third party

Census and vendor figures never blend into one number. Vendor attribution prints as a floor, never as an inference. No material non-public information is used or accepted.

Permitted field and gathering compression, ≥200 HP, excluding processing-plant and transmission compression, reconciled bottom-up where public disclosure allows.

Every covered basin, every quarter.

One basin publishes each Wednesday, majors first, until the rotation is complete. Then it runs again on refreshed counts, so every quarter you get the current picture of each basin and the moves that changed it.

Permian (TX + NM)

Gulf Coast / Eagle Ford

Marcellus / Utica

Mid-Continent / Anadarko

Powder River

Bakken

DJ Basin

East Texas / Haynesville

Rockies and legacy gas

The lower 48 · every unitLIVE FROM THE CENSUS

operator-ownedassigned vendorsized by installed HP · illustrative

Who reads it.

For compression providers

You know your own book. The count shows you the shape of the basin around it: who is setting, who owns against who rents, which operators run four-vendor books and which run none, and where the horsepower is migrating.

See the vendor read →

For analysts & investors

Vendor filings give you national fleet and utilization. They do not give you basin-level penetration, customer concentration by operator, or the standing rental layer underneath a record that reads owner-operated. The count estimates all three, with confidence intervals and stated method.

See the financial read →

For operators

You are negotiating renewals against a vendor who knows their fleet, their utilization, and their queue, and knows you do not. The count closes that gap: what your peers set, who serves them, how old the iron is, and when the paper around you rolls.

See the operator read →

The count is free because it is the proof. Every set, operator, engine, vintage, and vendor attribution the record supports, published whole, with no gate on the numbers.

The rate work, the forward paths, and the unit-level tables underneath are the business.

See Access →See The Census →

Bring us your basin.

Thirty minutes, your fleet or your counterparty list against the census, and you will know what the public regulatory record says about your position before you next sit down with a vendor.

Book a working session See the whole publication